Pseudo-work

Today I came across a social media post poking fun at the folks setting up their complex AI notetaking and note-analzying workflows. Hook up Email to Obsidian to OpenClawd, those sort of things. The post asked, are you really that important and busy that you can't go through your own emails and need a super-complex agent setup to do that for you?

It reminded me of Cal Newport's "Pseudo-productivity" idea, where visible activity is misunderstood as productivity even though real outcomes are lacking. In a similar vein, setting up a Rube Goldberg machine is fun. It feels like you're productive. You're building something real, maybe even putting some decent engineering effort into it. But at the end of the day, no effort should have been put into it:

• For simple tasks, pseudowork is overkill. It's procrastination disguised as building stuff.

• For hard tasks, pseudowork distracts. Hard tasks stay hard.

It comes back to identifying the true bottlenecks in any value stream. Sure, along that stream there might be steps that are more annoying than they need to be, but the real effort should be spent on the bottleneck, not on whatever can be automated in the most complicated way.